Pedogenic calcium carbonate (CaCO3) in soil in the arid and semi-arid area has become increasingly important as an indicator of plant nutrition, landscape age, paleoecology, and paleoclimate. The formation of CaCO3 in arid soils, however, is a debating problem because it is considered as either a biotic or an abiotic process. In the biotic process, the phenomenon of biomineralization has been reported by many scientists, while little direct attention has been given to the role of soil organic matter (OM) in the formation of pedogenic carbonate. The primary objective of this study is to demonstrate whether the OM is directly involved in the formation of pedogenic carbonate. We examined the relationships between OM and CaCO3 in Artemisia ordosica community on semi-fixed sandy land，Artemisia ordosica + Artemisia frigida community on fixed sandy land and Oxytropis aciphylla community on light brown soil land located on the eastern edge of the Tengri Desert in the League of Alashan, Inner Mongolia, China. Each of the three sampling plots is 80 m×80 m, each of which contained 400 quadrats with 4 m×4 m, respectively. Soil samples were taken from the upper 0~5 cm layer for each quadrat. The samples were sieved through a 0.45 mm sieve after air-dried. The OM and CaCO3 content were determined by the potassium dichromate volumetric and aerometric methods, respectively. Three replicates were conducted for each sample, and the mean value was used for the final analysis. The index C = s2/ū (ū is the mean value, s2: the variance) (distribution coefficient method), with the following interpretations: when C < 1, the pattern is uniform; C = 1: a random distribution and C > 1: a patchy distribution, was applied to determine the spatial distribution patterns of OM and CaCO3.Microsoft Excel and SPSS were used to analyze the regression relationships between OM and CaCO3 contents, and Geostatistics method was used to study the spatial heterogeneities of OM and CaCO3 based on the normalized data, so it is comparable to the spatial heterogeneities of different regionalized variables. The major findings were summarized as follows: (1) The distributions of CaCO3 and OM have a similar pattern to some extent. According to the results of distribution coefficient method, the OM had uniform distributions in all the three plots; CaCO3 had a random distribution in A. ordosica+A. frigida community, and uniform distributions in both A. ordosica and Oxytropis aciphylla community. The major reason is that the plants have more patchy distribution in A. ordosica+A. frigida community than in other two communities. (2) The content of CaCO3 was positively correlated with the content of OM, indicating that CaCO3 contents increased with the increase in OM contents. The OM content increased from A. ordosica to A. ordosica+A. frigida community, but decreased from A. ordosica+A. frigida to Oxytropis aciphylla community, so did the CaCO3 content. The results suggest that the OM in soil is greatly related to the pedogenic calcium carbonate. (3) To spatial heterogeneities, we analyzed the changes of sills, effective ranges, C/(C0+C)s and fractal dimensions of OM and CaCO3 in three sampling plots. The results showed that the effective ranges of OM and CaCO3 decreased gradually, indicating that the homogeneity ranges of OM and CaCO3 increased with the dune stabilization. The increase in fractal dimension of OM implied that the OM distribution tends to be more homogeneous. There was no obvious change in fractal dimension of CaCO3 in the process of dune stabilization, indicating that the formation of CaCO3 lagged the decomposition of OM. The increases in C/(C0+C)s of OM and CaCO3 suggested that their spatial autocorrelation increased in the process of dune stabilization. In general, the spatial heterogeneities of OM and CaCO3 were concordant in the three sampling plots. Therefore, the OM and CaCO3 were correlated with each other; the OM may be directly related to the formation of pedogenic carbonate in the process of dune stabilization.